Actuarial learning – how data science methods expand the actuary's toolbox
Valkyrja talk by Katrien Antonio
Valkyrja talk by Katrien Antonio at Reykjavik University March 8 in room V102 at 16:00.
The talk will start with an exploration of insurance pricing with statistical and machine learning methods. I will structure this exploration along four principles that are fundamental to actuarial learning methods for insurance pricing: (1) the inverted production cycle of the insurance business, (2) the applicable target variables, where frequency and severity variables are key, (3) the concept of pricing via risk classification and the need for smart feature engineering and feature selection in a highly regulated industry (4). I will then focus on our recent work on usage based insurance pricing with telematics collected data.
More info via https://katrienantonio.github.io
